Lessons learned from GMOs in regulating gene editing

It’s been 25 years since the first GMO crop was planted. And while GMOs have been a hot topic of debate over the last quarter century, this innovation has largely been celebrated by the agricultural community for its environmental benefits, its contribution to the farmers’ livelihoods and the economy, and for its contributions to advancing food security.

GMOs will continue to be an important tool for farmers, but there are new plant breeding techniques – gene editing being the most often talked about one – that are offering ever-expanding possibilities for developing new crops that meet the needs of farmers, consumers and the environment.

As governments look at how they are going to regulate products that come from these new breeding techniques, we can take away some key lessons from the regulation of GMOs over the last 25 years.

Back in 1996, the government of Canada deemed crops developed through recombinant DNA (rDNA) technology (or GMOs) are no more inherently risky than crop varieties developed through conventional breeding. As a result, they said that the regulation of plant breeding should focus on the characteristics, or “novelty,” of the final plant product, not the breeding process used to develop the plant. This is what is called a product-based approach to regulating plant breeding.

At the time, most of the rest of the world went on to adopt process-based approaches to regulating GMOs. This is to say, the use of rDNA technology itself automatically triggered regulatory oversight. Canada’s approach was largely viewed by experts and innovators as more scientifically sound and appropriate given the very low level of risk involved. The evidence we’ve seen over the last 25 years has confirmed the safety of GMOs and the lack of inherent risk that any individual breeding technique in and of itself presents. Canada clearly got it right in developing a product-based framework.

But a broader view of the last 25 years shows us Canada’s implementation of the product-based approach to regulating plant breeding has not actually been all that different to the process-based approach taken by much of the rest of the world. The types and number of products assessed (over 100 GMOs), and their respective safety assessments, have largely been the same.

The Canadian approach generally defines novelty as the presence of a characteristic significantly outside the anticipated range for the species. Although breeders self-assess novelty, and whether or not pre-market assessment applies to a product, the definition has proved challenging. There isn’t always a clearly established range for individual characteristics within a plant, and the whole point of plant breeding is to expand that range, which makes determining the goalposts for what is considered normal a subjective exercise.

As a result of the difficulty of providing clear guidance in a product-based system, all products of rDNA have required a pre-market assessment, and Canada has been the only country in the world to require additional safety assessments for products of conventional breeding.

Not only has this has led to confusion and uncertainty among innovators over what the requirements are for a product-based system and how they differ from process-based systems, it has also created an unintentional roadblock to innovation for products of conventional breeding, with no additional benefit for Canadians in terms of safety.

Furthermore, the model Canada uses to assess risk is the same one the rest of the world uses, which follows international standards for conducting safety assessments of food derived from rDNA plants[1]. The foundation for these risk assessments is a substantial equivalence comparison to a familiar/conventional product.

While the assessment approach is based on sound scientific principles, the issue with applying this approach in a product-based regulatory system is that this guidance has an underlying concern with the process of rDNA built in. That is, confirming the breeding process has not in and of itself negatively impacted the plant in some unforeseen way (e.g. created unintended effects), often makes up the bulk of the assessment and is the most labour-intensive part of the submission (including extensive field testing of agronomic and compositional characteristics). Moreover, these studies have never detected any meaningful differences.

Applying this part of the risk assessment is at odds with the founding principles of a product-based system and undermines the long track record of safety of conventionally bred plants and plant breeders’ ability to successfully manage unintended effects. Unintended effects are a natural part of plant breeding and the government is generally not concerned with their occurrence – thousands of new conventional varieties have been commercialized over the years without a novelty pre-market risk assessment, with no safety issues. The novelty of a specific plant characteristics does not impact the occurrence of these effects in plants, and therefore there is no reason to treat novel plant products any differently in these regards.

In a true product-based system the only point of interest should be in the known intended change. You can’t have a product-based trigger for regulatory oversight and then focus the risk assessment on process-based concerns. To be truly product-based, the assessment should focus on assessing the risks of the known novel trait on food, feed or environmental safety, not on the impacts the breeding process may have on the plant characteristics.

This assessment approach starts with the assumption that the new trait is expressed as claimed with no other negative unintended effects, and studies are not required to be submitted to prove it. This alone would make navigating the regulatory process much less burdensome, have significant positive impacts on innovation, with no compromise to safety. This assumption is supported by the thousands of conventional and over one hundred biotech varieties that have entered the market in the past decades with no unintended effects founds.

With new technology there may be questions about how the breeding process itself might impact the plant. In these situations, it may make sense for the regulator to ask for additional information, until they can be confident that the process itself doesn’t represent a concern.

This kind of additional information should be viewed as separate from that required as part of a product-based system. This should only be required for truly new technology with significant uncertainties that have not yet been thoroughly investigated. For gene editing specifically, the technology has been extensively studied for over a decade and the evidence shows that the chances for unintended effects to arise is no greater, and actually far less, than it is in conventional breeding.

To that end, Canadian innovators are looking for a modernized regulatory oversight program for plant breeding innovations. Designing a product-based system that is science-based, proportionate to risk, supports innovation and is predictable is not easy, but we believe it can be done. Canada is fortunate in that it has the framework in place for a product-based regulatory program that will ensure the health and safety of Canadians and can at the same time enable innovation. But the devil is in the details – and in this case, the implementation.

As new breeding techniques are rapidly emerging, Canada stands to be a model for the rest of the word in terms of how to regulate products of these new technologies. The government has committed to modernizing the regulatory system for plant breeding innovations and a key part of doing this is defining a truly product-based approach to regulation that provides clarity around novelty and is commensurate with real risk.

Jennifer Hubert, director, plant biotechnology, CropLife Canada

[1] Guideline for the conduct of food safety assessment of foods derived from recombinant-DNA plants